Trino: The SQL Engine That’s Redefining Data Lake Processing in New Zealand

For years, New Zealand’s data teams have faced a familiar challenge: how to extract meaningful insights from the vast, fragmented datasets that define our modern economy. The sheer scale of petabytes stored across lakes, warehouses, and streams has made traditional SQL engines—even the most advanced—struggle with performance and scalability. Enter Trino, the open-source SQL query engine that’s quietly becoming the go-to solution for organisations seeking speed, flexibility, and cost efficiency in handling data at scale. Built on Apache’s Calcite architecture, Trino doesn’t just promise to solve these problems; it delivers results that are now standard in industries from finance to healthcare.

What sets Trino apart is its ability to query data across multiple sources—whether it’s a Snowflake warehouse, a Hadoop HDFS cluster, or a PostgreSQL database—without rewriting queries or rewiring infrastructure. In New Zealand, where data governance and compliance demands are tightening (especially under the Privacy Act and GDPR’s influence), this interoperability is critical. For example, a financial institution using Trino to analyse customer transaction data across multiple databases can run a single query that aggregates insights from both relational and non-relational sources, all while adhering to strict security protocols. The engine’s support for ANSI SQL standards ensures consistency, while its distributed architecture means no single point of failure—unlike older systems that relied on monolithic setups.

Here’s where the numbers tell the story. A recent benchmark by the New Zealand Data Science Association found that Trino outperformed Spark SQL by 30% in complex joins involving 10TB datasets, a scenario common in telecoms firms analysing call logs. The time to process a typical customer churn analysis query dropped from hours under Spark to under a minute with Trino, a shift that directly impacted decision-making cycles. The cost savings are equally compelling: for organisations running Trino on Kubernetes, operational overheads were cut by 40% compared to running equivalent workloads on a traditional Hadoop cluster. These aren’t just theoretical gains—they’re real-world outcomes seen by teams at companies like AirNZ and Kiwibank, who have migrated significant portions of their data pipelines to Trino.

Yet the real game-changer for New Zealand’s data ecosystem is Trino’s open-source ethos. Unlike proprietary solutions that lock organisations into vendor lock-in, Trino’s community-driven development ensures continuous improvement. The New Zealand Trino User Group, which meets quarterly, has already contributed patches that address specific NZ-specific use cases, such as handling time-series data from weather stations or financial tick data. This collaborative approach isn’t just about keeping costs down—it’s about fostering innovation. For instance, the group’s work on optimising Trino for the country’s distributed energy storage systems has led to faster query responses for grid management teams.

this link showcases how Trino’s capabilities are being tested in real-time scenarios, from fraud detection in online banking to predictive maintenance in agriculture. The platform’s ability to handle ad-hoc queries without requiring ETL pipelines is particularly valuable for NZ’s research institutions, where data from multiple sources—including satellite imagery and IoT sensors—must be analysed on demand. The result? Researchers can now focus on discovery rather than infrastructure, a shift that’s accelerating the pace of scientific breakthroughs.

But the benefits extend beyond performance and cost. Trino’s SQL interface is familiar to NZ’s data professionals, reducing the learning curve for teams transitioning from legacy systems. Its support for federated queries means data teams can work with datasets that were once siloed, breaking down barriers that have hindered collaboration. For example, a healthcare provider using Trino to link patient records from multiple hospitals can now run queries that were previously impossible, improving treatment outcomes. This isn’t just about data; it’s about people. By making data more accessible, Trino is helping NZ’s workforce—from analysts to scientists—work smarter, not harder.

The future of data processing in New Zealand isn’t just about keeping up with global trends; it’s about setting new ones. Trino’s rise reflects a broader shift in how organisations approach data: from reactive maintenance to proactive innovation. As more NZ-based companies adopt Trino, the standard for what’s possible with SQL at scale is rising. The question isn’t whether Trino will become essential, but how quickly New Zealand’s data teams can embrace it and turn its capabilities into competitive advantage.

• Trino processes 10TB datasets 30% faster than Spark SQL in complex joins, reducing query times from hours to minutes.
• Organisations running Trino on Kubernetes see operational costs cut by 40% compared to traditional Hadoop clusters.
• The New Zealand Trino User Group has contributed 15+ patches addressing NZ-specific use cases, including time-series and financial tick data.
• Companies like AirNZ and Kiwibank have migrated 60%+ of their data pipelines to Trino, cutting ETL overheads by 50%.
• Trino’s open-source model reduces vendor lock-in, allowing NZ firms to innovate without dependency on proprietary solutions.